I read with this perspective with interest. I agree wholeheartedly that five years should be the norm for a
grant. It is important to note that there should be no need to flesh out the research for years 4 and 5 (as currently for a programme grant); results generate ideas, and I never knew a scientist who could not make good use of the extra time. One merit is that the number of applications would be
reduced dramatically; once steady state had been reached, there would be little financial impact. The reduced number of applications should ease the burden for the research councils and the reviewers because fewer applications in total would be needed to keep a project going over time (there would be, say, three years before planning the follow-up grant,whereas at present one starts the process when just one year into a grant).

Other merits of the system would be stability in funding for supporting the research of 4 year-PhD students and it would allow postdocs to be trained and brought up to speed before they have to leave (this is a problem many institutions, where we cannot usually recruit top-flight postdocs and spend up to two years waiting for the first good results). Postdocs might welcome the increased stability too.

I agree about the length of case, too. I have lobbied for some while for applications to be limited to 2 pages. One used to write 1.5-2 pages in applications for a PhD studentship, after all. This is sufficient material for a reviewer to judge whether there are any good and innovative ideas in the application. Too often the intellectual or creative basis of science is
drowned by data. On this topic, I submit that preliminary data is unnecessary from established scientists; their record will show whether they are competent to do experiments and to pick up new techniques and resources. Any preliminary data will have been gathered at someone's expense; possibly a postdoc employed on one grant has to work off target to gather data for the next grant. Aren't funding bodies colluding in misappropriation by demanding preliminary data when the phenomenon is established and the methods available? Should we write the paper before applying for funds?

Finally, I suggest that groups funded by research councils should be small. We are trained and judged by our abilities at the bench; we are not selected or appointed on the basis of our managerial skills. Every mature scientist not at the bench is wasting their training.